LLM Tracing
Every model call and tool call your AI agents make — latency, tokens, finish reason and errors, newest first.
Included with Smoo LLM Tracing - Starter, Smoo LLM Tracing - Pro and 1 more plansSee plans
LLM Tracing is the flight recorder for your AI agents. Each row is one turn — which model answered, how long it took, how many tokens it used, and whether it errored — coloured by duration so a slow or failed turn surfaces without being sorted for.
Cost is shown per turn where we can establish it — read off the call itself, or matched to its gateway spend record. A turn that went another route reads "not measured" rather than showing you a zero, and the footer tells you how many of the turns on screen are covered. Token counts arrive the same way and can be missing on the same turns.
How to
- 1
Pick a time window — 15m to 7d.
- 2
Filter by status, service or model to isolate a problem.
Status "error" plus the last hour is the fastest "why did the agent fail" view.
- 3
Paste a conversation ID to replay what the agent did in one conversation, turn by turn.